All search strategies were fully documented, including the date
of the search, limits applied and the number of references returned
by the search. Search results were imported into EndNote
(Thomson Reuters, New York) and copies of the libraries created
were stored on file. For each search, the initial number of search
results, the number of duplicates removed and the remaining
references to screen were recorded on search results forms. For
each reference, titles and abstracts were screened and any nonrelevant
publications eliminated, possible reasons for exclusions of references from ePlantLIBRA are: (a) Insufficient documentation
for evaluation; (b) Non-target area for example clinical areas different
from the selected ones or papers from other disciplines such
as plant science; (c) Review articles; (d) No control included in trial
(non-RCT) – Bioeffects papers; (e) Commentaries to the Editor; (f)
Unacceptable experimental/analytical procedures; (g)
Compositional data expressed in graphical/picture format (h)
Development of analytical methods for identification only; (i) no
analytical data.